摘要
为了实现污水处理过程中BOD的在线检测,提出了具有权函数动量项和自适应学习速率权值的新的过程神经网络改进算法,并利用该算法建立了BOD软测量模型,设计了基于ATMEGA1280单片机的系统主控制器,采用模块化软件设计理念,利用AVR Studio开发环境,编程实现数据采集、软测量、液晶显示、数据存储、打印驱动等功能。此仪表已在工业现场得到应用,平均相对预测误差小于4.1%。
For the online BOD monitoring in sewage-treatment process,an improved process neural network algorithm including function momentum adjustment item and learning rate automatic adjustment was introduced to establish the BOD soft sensing model.The main controller based on ATMEGA1280 single-chip was designed,the software was programmed by using the modularized software method and the integrated development environment of AVR Studio.The online BOD monitor had functions such as data acquisition,soft sensing,LCD display,data storing and printing,etc.The applied result indicates the average relative forecasting error of less than 4.1%.
出处
《化工自动化及仪表》
CAS
北大核心
2010年第4期61-63,85,共4页
Control and Instruments in Chemical Industry
基金
北京市自然科学基金资助项目(4062011)
北京市高校拔尖创新人才计划项目(200589)
北京市优秀人才项目
北京市教育委员会科技发展项目(KM200310011040)
